Accepted: 6 March 2023

DOI: 10.1111/1471-0528.17471

RESEARCH ARTICLE

Impact of coronavirus disease 2019 (COVID-­19) vaccination on

menstrual bleeding quantity: An observational cohort study

Blair G. Darney1,2   | Emily R. Boniface1   | Agathe Van Lamsweerde3  | Leo Han1  |

Kristen A. Matteson4  | Sharon Cameron5   | Victoria Male6   | Juan Acuna7  |
Eleonora Benhar3  | Jack T. Pearson3  | Alison Edelman1

1
Department of Obstetrics & Gynecology,
Oregon Heath & Science University, Portland,
Oregon, USA
2
Centro de Investigacion en Salud Poblacional
(CISP), Insituto Nacional de Salud Publica
(INSP), Cuernavaca, Mexico
3
Natural Cycles°, New York, NY, USA
4
University of Massachusetts Chan Medical
School, Worcester, Massachusetts, USA
5
Department of Obstetrics & Gynecology,
University of Edinburgh and Chalmers
Centre, NHS Lothian, Edinburgh, Scotland,
UK
6
Department of Metabolism, Digestion and
Reproduction, Imperial College London,
London, UK
7
Florida International University School of
Public Health, Miami, Florida, USA
Correspondence
Blair G. Darney, 3181 SW Sam Jackson Park
Road, UHN-­50, Portland, OR 97239, USA.
Email: darneyb@ohsu.edu
Funding information
National Institues of Health, Grant/Award
Number: NICHD089957
Abstract
Objective: To assess whether coronavirus disease 2019 (COVID-­19) vaccination im-
pacts menstrual bleeding quantity.
Design: Retrospective cohort.
Setting: Five global regions.
Population: Vaccinated and unvaccinated individuals with regular menstrual cycles
using the digital fertility-­awareness application Natural Cycles°.
Methods: We used prospectively collected menstrual cycle data, multivariable lon-
gitudinal Poisson generalised estimating equation (GEE) models and multivariable
multinomial logistic regression models to calculate the adjusted difference between
vaccination groups. All regression models were adjusted for confounding factors.
Main outcome measures: The mean number of heavy bleeding days (fewer, no
change or more) and changes in bleeding quantity (less, no change or more) at three
time points (first dose, second dose and post-­exposure menses).
Results: We included 9555 individuals (7401 vaccinated and 2154 unvaccinated).
About two-­t hirds of individuals reported no change in the number of heavy bleeding
days, regardless of vaccination status. After adjusting for confounding factors, there
were no significant differences in the number of heavy bleeding days by vaccination
status. A larger proportion of vaccinated individuals experienced an increase in total
bleeding quantity (34.5% unvaccinated, 38.4% vaccinated; adjusted difference 4.0%,
99.2% CI 0.7%–­7.2%). This translates to an estimated 40 additional people per 1000
individuals with normal menstrual cycles who experience a greater total bleeding
quantity following the first vaccine dose' suffice. Differences resolved in the cycle
post-­exposure.
Conclusions: A small increase in the probability of greater total bleeding quantity
occurred following the first COVID-­19 vaccine dose, which resolved in the cycle
after the post-­vaccination cycle. The total number of heavy bleeding days did not
differ by vaccination status. Our findings can reassure the public that any changes
are small and transient.

K EY WOR DS
bleeding quantity, COVID-­19 vaccination, menstrual cycle, menstruation

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited and is not used for commercial purposes.
© 2023 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd.

BJOG. 2023;00:1–10.  wileyonlinelibrary.com/journal/bjo   |  1


|
2       DARNEY et al.
1   | I N T RODUC T ION
Menstrual cycles are considered a sign of overall health, a
‘vital sign’ according to the US National Institutes of Health
(NIH) and medical societies.1,2 Experiencing an unexpected
change in menstruation can cause concern and even alarm.
Public and media reports about a possible link between
coronavirus disease 2019 (COVID-­19) vaccination and men-
strual disturbances have highlighted the lack of evidence to
respond to such concerns.3,4 Menstrual outcomes were not
included in COVID-­19 vaccine trials,5–­8 limiting the ability
of the manufacturers, public health agencies and clinicians
to respond to questions about the impact of the vaccine on
menstrual health. There are biologically plausible ways in
which a vaccine-­elicited immune response could cause men-
strual changes: cytokine production may transiently inter-
fere with the hypothalamic–­pituitary–­ovarian axis, which
drives the menstrual cycle,9–­12 and/or the activation of local
immune cells in the endometrium could impact tissue repair
at this site, potentially increasing menstrual bleeding.13,14
However, individuals naturally experience inherent and
normal variations in menstrual cycle duration and bleed-
ing patterns,15,16 making it challenging to isolate COVID-­19
vaccination as a cause.
A growing body of evidence demonstrates that COVID-­19
vaccination is associated with a small (less than 1 day) in-
crease in cycle length but with no change in the duration of
menses.17–­21 Other disturbances, such as bleeding quantity
or menstrual symptoms, are less well studied. Retrospective
studies have identified that changes in bleeding quantity may
occur with vaccination, but these studies are not designed to
determine whether vaccination is the main factor associated
with these changes, as they lack a comparison group and use
retrospective self-­reported data.22–­24
The objective of this study is to estimate the association
of COVID-­19 vaccination on menstrual bleeding quantity
among individuals with normal menstrual cycles (i.e. cycle
lengths of 24–­38 days with menses of ≤8 days). We examine
changes in the number of heavy bleeding days and in total
bleeding quantity using data from a retrospective cohort
study using prospectively collected menstrual cycle data and
an unvaccinated comparison group.
2   | M ET HODS
We conducted a retrospective cohort study using vaccination
and menstrual cycle data from individuals using the fertil-
ity awareness application Natural Cycles° (Natural Cycles°,
New York, NY, USA). Cycle data ranged from October 2020
to May 2022; initial COVID-­19 vaccinations were received
between January 2021 and April 2022. Individuals using
Natural Cycles° prospectively track their physiological
data as a non-­hormonal pregnancy prevention or planning
method, with no concurrent use of any hormonal contracep-
tive method. Details on variables tracked by Natural Cycles°
are reported elsewhere.15 Users may consent to the use of
14710528, 0, Downloaded from https://obgyn.onlinelibrary.wiley.com/doi/10.1111/1471-0528.17471 by INASP/HINARI - MONTENEGRO, Wiley Online Library on [14/04/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
their de-­identified data for research upon registering in the
app, and may remove their consent at any time. Individuals
who had consented to the use of their data for research pur-
poses (72% of all app users) were sent an in-­app message
between 7 May and 27 October 2021 explaining that their
data could be used for studies about the COVID-­19 vac-
cine and requesting their vaccination status (yes/no), and if
yes, vaccine dates (day, month, year) and vaccine brand. To
be eligible for study inclusion, individuals must have given
consent for the use of their anonymised data and reported
their vaccination information. We included individuals aged
18–­45 years who were at least three cycles post-­pregnancy
or from the use of hormonal contraception, were not meno-
pausal by self-­report and who had a normal pre-­vaccination
menstrual cycle (average length of 24–­38 days and menses
duration of ≤8 days).
Every individual contributed menses data from a mini-
mum of four consecutive cycles. For vaccinated individuals,
we included the three menses immediately prior to vaccina-
tion, and at least the menses associated with the first vac-
cine dose (designated as ‘first-­dose menses’). If available, we
also included menses data from the cycles associated with
the second vaccine dose (‘second-­dose menses’), as well as
the cycle and menses following vaccination (‘post-­exposure
menses’) to assess the potential resolution of changes. If an
individual was vaccinated during their menses, that menses
was designated as the ‘first-­dose menses’ or ‘second-­dose
menses’, respectively; if they were vaccinated after the com-
pletion of their menses, the following menses was designated
as the ‘first-­or second-­dose menses’. For unvaccinated in-
dividuals (control group), we included menses data from
four to six cycles from a similar time period, depending on
the volume of data recorded, to serve as the notional pre-­
vaccination period, first-­dose menses, second-­dose menses
and post-­exposure menses. Individuals without data beyond
menses associated with the first vaccination dose were ex-
cluded from the analyses of later time points.
Individuals using the Natural Cycles° application can
choose whether or not they want to track their menstrual
‘flow’ or menstrual bleeding quantity; it is not a required
variable and thus many individuals have no or incomplete
data for menstrual bleeding quantity. We excluded individ-
uals with more than 1 day of missing bleeding quantity data
during the three pre-­vaccination menses or during the three
post-­vaccination menses of interest. Thus, in total, individ-
uals were allowed up to 4 days of missing bleeding quantity
data to be included in the final data set: one during the pre-­
vaccination period and one for each of the post-­vaccination
menses analysed.
Our primary exposure was COVID-­19 vaccination, as re-
ported by individuals using the Natural Cycles° application:
individuals recorded their vaccination date(s) or confirmed
their unvaccinated status. We also classified vaccinated in-
dividuals by the mechanism of action of the vaccine that
they received: mRNA (Pfizer-­BioNTech and Moderna), ad-
enovirus vector (Astrazeneca, Johnson & Johnson/Janssen,
Covishield and Sputnik) and inactivated virus (Covaxin,

COVID19 VACCINE AND MENSTRUAL BLEEDING QUANTITY
Sinopharm and Sinovac). One individual who received
the protein subunit-­mediated Novavax vaccine could not
be classified with the other vaccine mechanisms, and was
therefore grouped with individuals with unspecified vaccine
brand.
Individuals report their daily bleeding quantity as
‘spotting’, ‘light’, ‘medium’ or ‘heavy’. We assessed men-
strual bleeding quantity in two ways: the number of heavy
bleeding days and the total bleeding quantity, i.e. the ordi-
nal sum of bleeding scores with spotting scored as 1, light
scored as 2, medium scored as 3, and heavy scored as 4. For
each post-­vaccination menses (first-­dose menses, second-­
dose menses and post-­exposure menses), we calculated the
within-­individual change from the median of the three pre-­
vaccination menses. We then categorised the change in the
number of heavy bleeding days as fewer days, no change
or more days, and categorised the change in total bleeding
quantity as less bleeding, no change or more bleeding. We
chose to categorise these changes rather than analyse them
as numeric because the units of the total bleeding quantity
outcome are not interpretable, and no clinical meaning ex-
ists for a fractional unit for either outcome. We therefore had
two outcomes measured at three time points each (the first-­
dose menses, second-­dose menses and post-­exposure men-
ses): (1) the change in number of heavy bleeding days; and (2)
the change in total bleeding quantity. We used a Bonferroni-­
adjusted significance level of 0.008 to account for multiple
comparisons for these six total outcomes and thus report
99.2% confidence intervals. No relevant core outcome set ex-
ists or is in development to address menstrual changes, and
therefore none were used for these analyses.
We included sociodemographic information collected by
Natural Cycles° using in-­app messages. Individuals using
the app are required to input some demographic variables
(such as age and country of residence) and other data are op-
tional (such as body mass index, BMI), whereas additional
sociodemographic data are supplied voluntarily as part of
research requests and are thus not supplied by all app users.
As such, some sociodemographic variables have a large vol-
ume of non-­ignorable missing data (for details on ‘missing-
ness’, see Table  S1). As a result, we included ‘missing’ as a
category in multivariable analyses. We classified age at the
beginning of the first cycle: 18–­24, 25–­29, 30–­34, 35–­39 or
40–­44 years. Individuals reported their race and ethnicity
as Asian, black, Hispanic, Middle Eastern or North African,
Native Hawaiian or Pacific Islander, or white. As a result
of the small sample sizes, we collapsed race and ethnicity
categories into a binary variable for multivariable analyses.
We categorised global region as UK and Channel Islands,
Europe, USA and Canada, Australia and New Zealand or
other. Individuals from Sweden made up the majority of
users in Europe (51%), whereas individuals from Brazil made
up the majority of users in the ‘other’ category (62%). We
grouped BMI into underweight, normal weight, overweight
and obese categories, and collapsed the underweight and
normal weight groups for multivariable analyses because of
the small sample sizes. We also included parity (nulliparous
14710528, 0, Downloaded from https://obgyn.onlinelibrary.wiley.com/doi/10.1111/1471-0528.17471 by INASP/HINARI - MONTENEGRO, Wiley Online Library on [14/04/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
  
   3|
vs parous), education (an undergraduate or higher degree vs
less education) and relationship status (in a steady relation-
ship or not).
The Oregon Health & Science University Institutional
Review Board (study no. 00023204, approved 6 August 2021)
and the UK's Reading Independent Ethics Committee (study
no. 230721, approved 23 July 2021) approved the study pro-
tocol. De-­identified data were used under a data use agree-
ment with Natural Cycles°. No members of the public were
directly involved in this analysis, although the research was
developed in response to public reports of menstrual distur-
bances following COVID-­19 vaccination.
2.1  | Analysis
We examined sociodemographic characteristics of the study
sample, by vaccination status and overall. We compared
the changes in the number of heavy bleeding days and total
bleeding quantity for the first-­dose, second-­dose and post-­
exposure menses using Pearson's chi-­square test, adjusting
the p-­values to reflect our significance level of 0.008. We
then created histograms of the raw, uncategorised change
in number of heavy bleeding days and total bleeding quan-
tity by vaccination group and tabulated the change in heavy
bleeding days. We developed multivariable longitudinal
Poisson general estimating equation (GEE) models for all
three heavy bleeding day outcomes. GEE models included
an offset for duration of menses, and an interaction between
time (pre-­/post-­vaccination) and vaccination status to deter-
mine the effect of vaccination, i.e. the adjusted difference in
the change in number of heavy bleeding days between vac-
cination groups. We then plotted the predicted number of
heavy bleeding days before and after vaccination for both
groups. For the three total bleeding quantity outcomes, we
developed multivariable multinomial logistic regression
models using no change as the base outcome, then calcu-
lated the adjusted predicted probability of each outcome
(less bleeding, no change or more bleeding) for each vaccina-
tion group and the adjusted difference between groups. All
regression models were adjusted for age, race and ethnicity,
parity, BMI, education, relationship status and global region.
Models for the second-­dose menses and post-­exposure men-
ses were also adjusted for time between the first and second
vaccine doses.
We conducted nine sensitivity analyses to confirm our
results. First, we compared the sociodemographic charac-
teristics of individuals included in the study with those who
were excluded for missing bleeding quantity data. Second,
we compared the change in the number of heavy bleeding
days and total bleeding quantity for the first-­and second-­
dose menses by vaccine mechanism of action (mRNA,
adenovirus vector or inactivated virus), by type of mRNA
vaccine and by timing of vaccination (during menses vs
after menses). Third, although the data did not meet the
missing-­at-­random assumption required for imputation
techniques, we completed 500 iterations of imputation

|
4       DARNEY et al.
and weighting with covariate balancing propensity scores
using bootstrapped standard errors to confirm that our
findings were not biased by missing data or by differences
in the characteristics of the vaccination groups. Fourth, we
excluded any individuals with polycystic ovary syndrome
(PCOS), endometriosis or thyroid disorder (n  =  454).
Fifth, we excluded any individual who used emergency
contraception during the study period (n = 472). Sixth, we
excluded individuals with any pre-­vaccination cycles out-
side the normal length of 24–­38 days (n = 1420). Seventh,
we excluded individuals who had received both vaccine
doses since their previous menses (n  =  422). Eighth, we
excluded individuals with any missing bleeding quan-
tity data during the pre-­ vaccination menses, or first-­
dose, second-­dose or post-­exposure menses (n = 4090 for
first-­dose menses, n  =  3050 for second-­dose menses and
n = 3020 for post-­exposure menses). Finally, we included
individuals with pre-­vaccination menses durations of 9 or
10 days (who had previously been excluded; n  =  93). We
used Stata 15.1 (StataCorp LLC, College Station, TX, USA)
for all analyses.
3   | R E SU LT S
Out of 42 095 users, 9555 individuals (7401 vaccinated and
2154 unvaccinated) representing 229 320 menses met the
inclusion criteria (Figure 1). The overall cohort was under
the age of 35 years (80.4%), nulliparous (77.4%), had at
least a college degree (67.8%) and was in a steady relation-
ship (69.7%) (Table  1). Most individuals were located in
the UK (32.4%), Europe (31.2%) or the USA and Canada
(30.1%). Among vaccinated individuals, the majority re-
ceived mRNA vaccines: 66.7% Pfizer-­BioNTech and 17.9%
Moderna. Vaccinated individuals were more likely to
have at least an undergraduate degree (71.4%, compared
with 55.3% among unvaccinated individuals). The char-
acteristics of individuals included in the study were simi-
lar to those excluded for missing bleeding quantity data
(Table S2); the majority of participants excluded for miss-
ing bleeding quantity data were excluded for missing data
in the pre-­vaccination period (8851 out of 10 325 excluded,
86%; data not shown), and thus a baseline could not be
established. The sample size for the vaccinated group
was considerably smaller for the second-­ dose menses
(n = 5288; Figure 1) for a variety of reasons: some individu-
als (2.0% of vaccinated) received the single-­dose Johnson
& Johnson/Janssen vaccine; some did not receive or did
not record information about their second dose (22.6%);
and some (3.1%) received their second dose after their last
cycle of tracked data.
The unadjusted change in the number of heavy bleed-
ing days reported did not differ between the vaccinated and
unvaccinated control group for any of the post-­vaccination
menses (Figure S1; Tables 2 and S3). Regardless of vaccination
status (actual or notional) for the first-­dose menses, second-­
dose menses and post-­exposure menses, approximately 60%
14710528, 0, Downloaded from https://obgyn.onlinelibrary.wiley.com/doi/10.1111/1471-0528.17471 by INASP/HINARI - MONTENEGRO, Wiley Online Library on [14/04/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
of individuals experienced no change, whereas approxi-
mately 20% experienced fewer heavy bleeding days and 20%
experienced more heavy bleeding days. After adjustment in
multivariable models, there were still no significant differ-
ences by vaccination status during the first-­and second-­dose
menses (Figure 2).
When examining the change in total bleeding quan-
tity pre-­and post-­vaccination (less bleeding, no change
or more bleeding), we found differences between the
vaccinated and unvaccinated control groups (Figure  S2;
Table 2). Vaccinated individuals were more likely to report
more bleeding overall than their unvaccinated counter-
parts during both the first-­dose menses (38.3% vs 34.8%,
p  =  0.027) and the second-­dose menses (39.8% vs 35.5%,
p < 0.001). The proportion of individuals who experienced
no change in total bleeding quantity was roughly 18% for
both vaccinated and unvaccinated groups at both time
points. The differences between groups were no longer sta-
tistically significant in the post-­exposure cycle. After ad-
justment in multivariable models, the predicted probability
of experiencing more total bleeding quantity was higher
in the vaccinated group compared with the unvaccinated
control group: 4.0% higher (99.2%  CI 0.7%–­7.2%) for the
first-­dose menses and 3.8% higher (99.2%  CI 0.2%–­7.3%)
for the second-­dose menses (Table S4). This translates to
approximately 40 additional people per 1000 individuals
with normal cycles who will experience a greater total
bleeding quantity as a result of vaccination. Again, this
difference was no longer significant in the post-­exposure
menses: 2.8% higher for vaccinated individuals (99.2% CI
−0.8%–­6.3%; data not shown).
We found no major differences in the change in number
of heavy bleeding days or the total bleeding quantity when
comparing individuals who received an mRNA vaccine
with individuals who received an adenovirus vector vaccine
during either the first-­or second-­dose menses (Figures  S3
and S4), and there were no differences in any bleeding out-
come between the two mRNA vaccines (data not shown).
The number of individuals who received an inactivated virus
(n  =  44 for first-­dose menses and n  =  31 for second-­dose
menses) was too small to draw conclusions. We also found
no meaningful differences in our outcomes when examined
by timing of vaccination (during menses vs after completion
of menses; Figures S5 and S6).
The sensitivity analysis imputing missing data and
weighting the sample to balance vaccination groups on
all sociodemographic variables did not alter our unad-
justed or adjusted results in a meaningful way (Table S5).
Nor did the sensitivity analyses excluding individuals: (1)
with PCOS, endometriosis or thyroid disorders; (2) who
used emergency contraception; (3) with pre-­vaccination
cycle lengths outside the normal range; (4) who received
two vaccine doses prior to a single menses; or (5) with any
missing bleeding quantity data. The final sensitivity anal-
ysis including individuals with slightly longer durations
of pre-­vaccination menses (9–­10 days) also did not change
the results.
 14710528, 0, Downloaded from https://obgyn.onlinelibrary.wiley.com/doi/10.1111/1471-0528.17471 by INASP/HINARI - MONTENEGRO, Wiley Online Library on [14/04/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
COVID19 VACCINE AND MENSTRUAL BLEEDING QUANTITY   
   5|

Eligible (n= 42 095)

Identification

Not included (n= 32 540)

No logged cycle data (n= 332)
< 3 pre-vaccination cycles (n= 10 871)
No vaccination cycle data (n= 1061)
Pre-vaccination menses length > 10
days (n= 601)
Non-consecutive cycles (n = 2675)
< 3 cycles post-pregnancy (n= 1909)
< 3 cycles post-hormonal
Inclusion contraception (n= 2982)
Menopausal (n= 30)
Outside of age range (n= 2)
Average pre-vaccination cycle length
< 24 or > 38 days (n= 1659)
> 1 missing day of bleeding quantity
data for 3 pre-vaccination menses or
for post-vaccination outcome menses
(n= 10 325)
Pre-vaccination menses length of 9-10
days (n= 93)a

Included in dataset (n= 9555)

Reported COVID-19 Vaccination Reported unvaccinated for COVID-19

1st dose mensesb: n= 7401 Notional 1st dose menses: n= 2154
Analysis Notional 2nd dose menses: n= 2134
2nd dose mensesc: n= 5288
Post-exposure mensesd: n= Notional post-exposure menses: n=
5127 2106

F I G U R E 1   STROBE flow diagram. aIndividuals with pre-­vaccination menses of 9–­10 days in duration were excluded from all primary analyses but
were then later included for a sensitivity analysis. b1st dose menses indicates the menses during or immediately following receipt of the first COVID-­19
vaccine dose. c2nd dose menses indicates the menses during or immediately following receipt of the second COVID-­19 vaccine dose. dPost-­exposure
menses indicates the menses of the cycle following the second-­dose menses

4   | DISC US SION days tracked, regardless of vaccination status. After adjust-

4.1  |  Main findings
We used prospectively tracked menstrual cycle bleeding data
to assess changes in the number of heavy bleeding days and
total bleeding quantity (menstrual ‘flow’) in individuals who
received COVID-­19 vaccination, as compared with an un-
vaccinated control group. Overall, about two-­thirds of indi-
viduals reported no change in the number of heavy bleeding
ing for confounding factors, we found no significant differ-
ences in the number of heavy bleeding days by vaccination
status. However, a larger proportion of vaccinated individu-
als experienced an increase in total bleeding quantity. Of the
unvaccinated individuals, 34.5% experienced a greater bleed-
ing quantity following the first vaccine dose, compared with
38.4% among the vaccinated individuals: representing an ad-
justed increase of 4.0% (99.2% CI 0.7%–­7.2%) in the probabil-
ity of a greater bleeding quantity following the first vaccine
 14710528, 0, Downloaded from https://obgyn.onlinelibrary.wiley.com/doi/10.1111/1471-0528.17471 by INASP/HINARI - MONTENEGRO, Wiley Online Library on [14/04/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
6       DARNEY et al.

T A B L E 1   Study participant characteristics (n = 9555), by vaccination status

Characteristic Unvaccinated Vaccinated Overall p

n 2154 7401 9555
Age group (years)
18–­24 423 (19.6) 745 (10.1) 1168 (12.2) <0.001
25–­29 724 (33.6) 2571 (34.7) 3295 (34.5)
30–­34 611 (28.4) 2609 (35.3) 3220 (33.7)
35–­39 296 (13.7) 1109 (15.0) 1405 (14.7)
40–­4 4 100 (4.6) 367 (5.0) 467 (4.9)
BMI category
Underweight 52 (2.4) 196 (2.7) 248 (2.6) 0.004
Normal weight 951 (44.2) 3468 (46.9) 4419 (46.3)
Overweight 277 (12.9) 993 (13.4) 1270 (13.3)
Obese 123 (5.7) 486 (6.6) 609 (6.4)
No data 751 (34.9) 2258 (30.5) 3009 (31.5)
Race and ethnicity
Asian 7 (0.3) 73 (1.0) 80 (0.8) <0.001
Black 46 (2.1) 54 (0.7) 100 (1.1)
Hispanic 47 (2.2) 145 (2.0) 192 (2.0)
Middle Eastern or North African 7 (0.3) 19 (0.3) 26 (0.3)
Native Hawaiian or Pacific Islander 3 (0.1) 15 (0.2) 18 (0.2)
Non-­Hispanic white 829 (38.5) 2926 (39.5) 3755 (39.3)
No data 1215 (56.4) 4169 (56.3) 5384 (56.4)
Parity
Nulliparous 1513 (70.2) 5882 (79.5) 7395 (77.4) <0.001
Parous 332 (15.4) 824 (11.1) 1156 (12.1)
No data 309 (14.4) 695 (9.4) 1004 (10.5)
Education
Less than college degree 548 (25.4) 1093 (14.8) 1641 (17.2) <0.001
Completed college degree 1190 (55.3) 5284 (71.4) 6474 (67.8)
No data 416 (19.3) 1024 (13.8) 1440 (15.1)
Relationship status
Not in steady relationship 313 (14.5) 962 (13.0) 1275 (13.3) <0.001
In steady relationship 1423 (66.1) 5236 (70.8) 6659 (69.7)
No data 418 (19.4) 1203 (16.3) 1621 (17.0)
Geographic region
UK 521 (24.2) 2574 (34.8) 3095 (32.4) <0.001
Europe 519 (24.1) 2466 (33.3) 2985 (31.2)
USA and Canada 931 (43.2) 1941 (26.2) 2872 (30.1)
Australia and New Zealand 152 (7.1) 222 (3.0) 374 (3.9)
Other 31 (1.4) 198 (2.7) 229 (2.4)
Vaccine type
Pfizer 0 (0) 4938 (66.7) 4938 (51.7) N/Aa
Moderna 0 (0) 1322 (17.9) 1322 (13.8)
Johnson & Johnson 0 (0) 151 (2.0) 151 (1.6)
Astrazeneca/adenovirus vector 0 (0) 650 (8.8) 650 (6.8)
Whole/inactivated virus 0 (0) 44 (0.6) 44 (0.5)
Unspecifiedb 0 (0) 296 (4.0) 296 (3.1)
Unvaccinated 2154 (100) 0 (0) 2154 (22.5)

Note: p-­values represent differences by vaccination status using Pearson's chi-­square test.
a
No statistical test performed.
b
Unspecified group contains one individual who received the protein subunit-­based Novavax vaccine.
 14710528, 0, Downloaded from https://obgyn.onlinelibrary.wiley.com/doi/10.1111/1471-0528.17471 by INASP/HINARI - MONTENEGRO, Wiley Online Library on [14/04/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
COVID19 VACCINE AND MENSTRUAL BLEEDING QUANTITY    |
   7

T A B L E 2   Changes in number of heavy bleeding days from the median of three pre-­vaccination menses to the first-­dose menses, second-­dose menses
and post-­exposure menses, by vaccination status

Difference from pre-­

Outcome vaccination to: Category Unvaccinated Vaccinated pa
Number of heavy First-­dose mensesb n 2154 7401 1.000
bleeding days Fewer heavy days 420 (19.5) 1371 (18.5)
No change 1304 (60.5) 4630 (62.6)
More heavy days 430 (20.0) 1400 (18.9)
c
Second-­dose menses n 2134 5288 1.000
Fewer heavy days 390 (18.3) 916 (17.3)
No change 1299 (60.9) 3270 (61.8)
More heavy days 445 (20.9) 1102 (20.8)
d
Post-­exposure menses n 2106 5127 1.000
Fewer heavy days 400 (19.0) 908 (17.7)
No change 1282 (60.9) 3197 (62.4)
More heavy days 424 (20.1) 1022 (19.9)
Total bleeding First-­dose mensesb n 2154 7401 0.027
quantity Less quantity 1013 (47.0) 3208 (43.4)
No change 392 (18.2) 1356 (18.3)
More quantity 749 (34.8) 2837 (38.3)
Second-­dose mensesc n 2134 5288 <0.001
Less quantity 983 (46.1) 2161 (40.9)
No change 393 (18.4) 1020 (19.3)
More quantity 758 (35.5) 2107 (39.8)
Post-­exposure mensesd n 2106 5127 0.157
Less quantity 973 (46.2) 2192 (42.8)
No change 369 (17.5) 969 (18.9)
More quantity 764 (36.3) 1966 (38.4)

Note: Data are n (%).

a
p-­values are adjusted to account for six comparisons; values below 0.05 represent statistically significant differences.
b
First-­dose menses indicates the menses during or immediately following receipt of the first COVID-­19 vaccine dose.
c
Second-­dose menses indicates the menses during or immediately following receipt of the second COVID-­19 vaccine dose.
d
Post-­exposure menses indicates the menses of the cycle following the second-­dose menses.

dose in multivariable analyses. For the second dose, 35.9% of
the unvaccinated individuals compared with 39.7% of the vac-
cinated individuals experienced a greater bleeding quantity
following vaccination: a 3.8% higher (99.2%  CI 0.2%–­7.3%)
adjusted probability for the second-­dose menses in multivari-
able analyses. This translates into an estimated additional 40
people per 1000 individuals with normal cycles (the difference
between 342/1000 unvaccinated and 384/1000 vaccinated)
who will experience a greater total bleeding quantity with
the first vaccine dose as a result of vaccination. The bleeding
quantity recorded by vaccinated and unvaccinated groups no
longer differed by the time of the post-­exposure cycle.
4.2  |  Strengths and limitations
The strengths of this study include a large global sample with
geographic diversity and prospectively tracked bleeding data:
users tracked bleeding quantity each day (vs the retrospective
self-­reported perceptions of changes in bleeding that were previ-
ously reported). Our outcome measures used counts of heavy
bleeding days and an ordinal measure of total bleeding quantity,
which are less subjective than retrospective reports of ‘more’ or
‘less’ bleeding (as previously reported). We also include an un-
vaccinated control group that allows us to isolate the relation-
ship of vaccination and menstrual bleeding quantity.
The limitations of our study include a lack of complete
sociodemographic data to fully understand the racial/ethnic
and gender-­identity diversity in our sample. Our study popu-
lation is also of lower BMI and higher education than is found
in general populations. Further, our sample is global, but the
USA, UK and Europe are over-­represented. To isolate the
association of the COVID-­19 vaccine and menstrual bleed-
ing changes, our study population includes individuals with
normal menstrual characteristics pre-­vaccine. Future studies
should focus on key subgroups, including contraceptive users,
menopausal individuals, people using gender-­affirming hor-
monal therapy and those with non-­clinically normal cycles
 14710528, 0, Downloaded from https://obgyn.onlinelibrary.wiley.com/doi/10.1111/1471-0528.17471 by INASP/HINARI - MONTENEGRO, Wiley Online Library on [14/04/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
8       DARNEY et al.

F I G U R E 2   Predicted number of heavy bleeding days for the median of three pre-­vaccination menses and the first-­dose (left), second-­dose (centre)
and post-­exposure (right) menses. Estimates are from longitudinal GEE models with an offset for menses duration, an interaction between vaccination
status and pre-­/post-­vaccination timing, and adjusted for age, BMI, educational attainment, parity, relationship status, global region and time between
doses (second dose and post-­exposure only). Unvaccinated users are shown in red, vaccinated users are shown in blue and error bars represent the
99.2% CIs. First-­and second-­dose menses indicate the menses during or immediately following receipt of the first and second vaccine dose, respectively.
Post-­exposure menses indicates the menses of the cycle following the second-­dose menses

or menses at baseline. Our sample may also be subject to se-
lection bias if individuals who experienced bleeding changes
were more likely to provide information on their vaccina-
tion status than individuals with no changes. However, the
existence of a sizable unvaccinated cohort suggests that our
study population is willing to contribute to menstrual cycle
research, regardless of the potential impact of the vaccine.
We recognise that more questions remain regarding dif-
ferent aspects of menstrual bleeding, but we are limited by
the variables collected within the tracking application. On
the one hand, the app utilises the gold standard for real-­time
menstrual cycle tracking and identifying heavy bleeding,
which is based on an individual's self-­assessment.16 On the
other hand, the units of our ordinal measure of total bleed-
ing quantity have no real numerical interpretation and are
not necessarily comparable with the self-­reported outcomes
of ‘more’ or ‘less’ bleeding reported previously. However, this
outcome enables us to note whether a change in bleeding
quantity or ‘menstrual flow’ has occurred or not in relation
to vaccine exposure, which is a critical knowledge gap and
an important outcome of documented interest to the public.
Additionally, individuals are not required to track bleeding
quantity data in the app, which limited our sample size. We
examined the ‘missingness’ of the data between included and
excluded individuals and found no changes in our results. We
have no measure of COVID-­19 disease, which may impact
bleeding patterns,25–­27 and vaccination status is self-­reported
and not verified, although we do have dates and vaccine
brands, and the self-­report of vaccine status and timing has
been shown to be a good surrogate for clinical data.28
4.3  |  Interpretation in light of other evidence
A study using the retrospective recall of changes in bleed-
ing quantity following COVID-­19 vaccination and no com-
parison group found that 42% of those reporting regular
menstrual cycles prior to vaccination reported experienc-
ing heavier bleeding following vaccination, relative to their
bleeding before vaccination.24 Another study also using
retrospective recall of menstrual changes but including
an unvaccinated pre-­ /post-­
comparison group (surveyed
individuals were asked about outcomes before and after
vaccination) found that menstrual disturbances did fol-
low vaccination. However, this study also identified that
more than one-­third of their total participants (vaccinated
and control) experienced disturbances even prior to vac-
cination or for controls over a similar time period.22 This
finding is consistent with our results that over one-­t hird of
individuals, regardless of vaccination status, had a greater
bleeding quantity. A retrospective study without an unvac-
cinated comparison group reported that menstrual changes

COVID19 VACCINE AND MENSTRUAL BLEEDING QUANTITY
were more frequent following the second dose than the first
dose;23 another found no differences in menstrual flow.18
Our study highlights the importance of an unvaccinated
comparator group: bleeding variability was high, regardless of
receiving a vaccine. With the inclusion of an unvaccinated con-
trol group and accounting for confounding factors, we found no
differences in the number of heavy bleeding days by vaccina-
tion status, but we did find that a small but significantly larger
proportion of vaccinated individuals experienced a greater
total bleeding quantity in the cycle of their first-­and second-­
vaccine doses, compared with unvaccinated individuals. This
difference was resolved by the time of the cycle following ex-
posure to COVID-­19 vaccination. Previous research using the
Natural Cycles° app data found a small increase in cycle length
and no difference in menses length associated with COVID-­19
vaccination.20,21 This study adds to the growing body of evi-
dence of time-­limited menstrual disturbances associated with
COVID-­19 vaccination at a population level of reproductive-­
aged individuals with previously regular menstrual cycles.
The impact of COVID-­19 disease on menstrual distur-
bances in unvaccinated individuals is less well defined, which
is the relevant counterfactual for those hesitant to vaccinate
themselves or family members because of concerns about
menstrual disturbances. Changes in menstrual cycles were
reported by 16% of a cohort who had COVID-­19 disease prior
to the availability of vaccines (2020), compared with those
who had COVID-­19 disease in 2020 and reported no change
in menstrual cycles. In two studies carried out prior to the
availability of vaccines, 16% or 28% of menstruating people
with COVID-­19 disease reported menstrual changes, with
changes associated with more severe disease.25, 26 In a cohort
of people with long COVID, more than 30% of menstruating
participants reported some kind of menstrual disturbance.27
Those who reported menstrual changes also reported more
severe COVID-­19 symptoms.25 Finally, COVID-­19 disease
carries serious morbidity and mortality risks to unvaccinated
individuals, which must be considered when discussing con-
cerns about menstrual disturbances and vaccination.
The development of core outcome sets and regulatory
support to include these core outcomes in future vaccine
development studies would aid in providing foundational
information on a critical patient-­reported outcome.
In conclusion, we find a small but statistically significant
increase (4.0%) in the adjusted probability of a greater total
bleeding quantity following the first COVID-­19 vaccine dose,
compared with an unvaccinated comparison group. This dif-
ference translates into an estimated additional 40 people per
1000 individuals with normal cycles experiencing a greater
bleeding quantity post-­vaccine that is resolved by the time of
the post-­vaccination cycle. We found no difference in the total
number of heavy bleeding days by vaccination status. Our
findings can inform patients and help providers in counsel-
ling about what to expect following a COVID-­19 vaccination.
AU T HOR C ON T R I BU T ION S
AE conceived of the study and secured funding. AE, ERB,
and BGD developed the analysis plan and conducted
14710528, 0, Downloaded from https://obgyn.onlinelibrary.wiley.com/doi/10.1111/1471-0528.17471 by INASP/HINARI - MONTENEGRO, Wiley Online Library on [14/04/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
  
   9|
analyses. BGD and ERB drafted the manuscript. AE, EB,
JTP, AVL, KAM, SC, LH, JA, and VM provided substantive
intellectual contributions to the manuscript.
AC K N O​W L E​D G E​M E N T S
We thank the individuals who shared their menstrual ex-
periences around the COVID-­19 vaccine, whose voices in-
formed our work. We also acknowledge the contribution to
science that would not have been possible without the users
of Natural Cycles who granted permission to share their de-­
identified data.
F U N DI N G I N F OR M AT ION
Research reported in this publication was externally peer-­
reviewed and funded by the Eunice Kennedy Shriver
National Institute of Child Health and Human Development
(NICHD) and the NIH Office of Research on Women’s
Health (NIH NICHD089957 Supplement). The funder
played no role in the study design, analysis or interpretation
of the data presented in this article.
C ON F L IC T OF I N T E R E S T S TAT E M E N T
AE reports honoraria and travel reimbursement from
American College of Obstetricians and Gynecologists
(ACOG), World Health Organization (WHO) and
Gynuity for committee activities, and honoraria for peer
review from the Karolinska Institute. AE receives roy-
alties from UpToDate, Inc. Oregon Health & Science
University (OHSU) receives research funding from the
OHSU Foundation, Merck, HRA Pharma and NIH, for
which AE is the principal investigator. BGD reports hon-
oraria and travel reimbursement from ACOG and the
Society for Family Planning (SFP) for board, committee
and mentorship activities. OHSU receives research fund-
ing from Merck/Organon and the Office of Population
Affairs (OPA)/Department of Health and Human
Services (DHHS), for which BGD is the principal inves-
tigator. OHSU receives research funding from the OHSU
Foundation, the Bill & Melinda Gates Foundation, the
American Board of Obstetrics and Gynecology (ABOG),
the American Society for Reproductive Medicine (ASRM)
and the NIH, for which LH is the principal investigator.
EB, AVL and JTP are employees of Natural Cycles°. KAM
reports honoraria and travel reimbursement from ABOG
and travel reimbursement from ACOG. The Women &
Infants Hospital received funding from Myovant for con-
sulting work performed by KAM on outcome measures for
heavy menstrual bleeding. ERB did not report any poten-
tial conflicts of interest. JA reports honoraria from Natural
Cycles. VM reports research funding from Borne, payment
for acting as an external examiner for the Universities of
Cambridge, Leeds, Swansea and Trinity College Dublin,
payment for articles written for the Guardian newspaper,
royalties received for contribution to Immunology 9th
edition (Elsevier) and support to attend the 16th Vaccine
Congress. Completed disclosure of interests form available
to view online as supporting information.

|
10       DARNEY et al.
DATA AVA I L A BI L I T Y S TAT E M E N T
The data that support the findings of this study are available
from NaturalCycles and were used under a data use agree-
ment. Restrictions apply to the availability of these data.
E T H IC S A PPROVA L
The OHSU Institutional Review Board approved the proto-
col. De-­identified data were used under a data-­use agreement
with Natural Cycles° and from the Reading Independent
Ethics Committee.
S U PP ORT I N G I N F OR M AT ION
Additional supporting information may be found in the on-
line version of the article at the publisher's website.
ORC I D
Blair G. Darney  https://orcid.org/0000-0001-8120-028X
Emily R. Boniface  https://orcid.org/0000-0001-8012-1758
Sharon Cameron  https://orcid.org/0000-0002-1168-2276
Victoria Male  https://orcid.org/0000-0001-5654-5083
R EFER ENCES
1. ACOG, Committee on Adolescent Health Care. Menstruation in girls
and adolescents: using the menstrual cycle as a vital sign. Committee
opinion #651. Obstet Gynecol. 2015;126:e143–­6.
2. Tingen CM, Halvorson LM, Bianchi DW. Revisiting menstru-
ation: the misery, mystery, and marvel. Am J Obstet Gynecol.
2020;223(5):617–­8.
3. Morgan E. Periods: why women's menstrual cycles have gone haywire.
The Guardian. 2021 Mar 25.
4. Landman K. Why were scientists so slow to study Covid-­19 vaccines and
menstruation? [Internet]. Vox. 2022 [cited 2022 Aug 1]. Available from:
https://www.vox.com/22935​125/covid​-­vacci​ne-­trial​s-­menst​rual-­c ycle​
-­perio​d-­chang​es-­ferti​lity-­myths
5. Polack FP, Thomas SJ, Kitchin N, Absalon J, Gurtman A, Lockhart S,
et al. Safety and efficacy of the BNT162b2 mRNA Covid-­19 vaccine.
N Engl J Med. 2020;383(27):2603–­15.
6. Oliver SE. The Advisory Committee on Immunization Practices' Interim
Recommendation for Use of Moderna COVID-­19 Vaccine –­United
States, December 2020. MMWR Morb Mortal Wkly Rep. 2021;69:1653–­6.
7. Sadoff J, Gray G, Vandebosch A, Cárdenas V, Shukarev G, Grinsztejn
B, et al. Safety and efficacy of single-­dose Ad26.COV2.S vaccine
against Covid-­19. N Engl J Med. 2021;384(23):2187–­201.
8. Baden LR, El Sahly HM, Essink B, Kotloff K, Frey S, Novak R, et al.
Efficacy and safety of the mRNA-­1273 SARS-­CoV-­2 vaccine. N Engl J
Med. 2021;384(5):403–­16.
9. Turnbull AV, Rivier CL. Regulation of the hypothalamic-­pituitary-­
adrenal axis by cytokines: actions and mechanisms of action. Physiol
Rev. 1999;79(1):1–­71.
10. Karagiannis A, Harsoulis F. Gonadal dysfunction in systemic dis-
eases. Eur J Endocrinol. 2005;152(4):501–­13.
11. Bergamaschi C, Terpos E, Rosati M, Angel M, Bear J, Stellas D, et al.
Systemic IL-­15, IFN-­γ, and IP-­10/CXCL10 signature associated with
effective immune response to SARS-­CoV-­2 in BNT162b2 mRNA vac-
cine recipients. Cell Rep. 2021;36(6):109504.
12. Teijaro JR, Farber DL. COVID-­19 vaccines: modes of immune activa-
tion and future challenges. Nat Rev Immunol. 2021 Apr;21(4):195–­7.
13. Lee KMN, Junkins EJ, Luo C, Chongliang L, Fatima UA, Cox
ML, et al. Investigating trends in those who experience men-
strual bleeding changes after SARS-­ CoV-­2 vaccination. Sci Adv.
2022;8(28):eabm7201.
14. Monin L, Whettlock EM, Male V. Immune responses in the human
female reproductive tract. Immunology. 2020;160(2):106–­15.
14710528, 0, Downloaded from https://obgyn.onlinelibrary.wiley.com/doi/10.1111/1471-0528.17471 by INASP/HINARI - MONTENEGRO, Wiley Online Library on [14/04/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
15. Bull JR, Rowland SP, Scherwitzl EB, Scherwitzl R, Danielsson KG,
Harper J. Real-­world menstrual cycle characteristics of more than
600,000 menstrual cycles. NPJ Digit Med. 2019;2:83.
16. Munro MG, Critchley HOD, Fraser IS, FIGO Menstrual Disorders
Committee. The two FIGO systems for normal and abnormal uterine
bleeding symptoms and classification of causes of abnormal uterine
bleeding in the reproductive years: 2018 revisions. Int J Gynaecol
Obstet. 2018;143(3):393–­408.
17. Wang S, Mortazavi J, Hart JE, Hankins JA, Katuska LM, Farland LV,
et al. A prospective study of the association between SARS-­CoV-­2 in-
fection and COVID-­19 vaccination with changes in usual menstrual
cycle characteristics. Am J Obstet Gynecol. 2022;227:739.e1–­11.
18. Alvergne A, Woon EV, Male V. Effect of COVID-­19 vaccination
on the timing and flow of menstrual periods in two cohorts. Front
Reprod Health. 2022;25(4):952976.
19. Gibson EA, Li H, Fruh V, Gabra M, Asokan G, Jukic AMZ,
et al. Covid-­19 vaccination and menstrual cycle length in the Apple
Women's Health Study. NPJ Digit Med. 2022;5(1):165.
20. Edelman A, Boniface ER, Benhar E, Han L, Matteson KA, Favaro C,
et al. Association between menstrual cycle length and coronavirus
disease 2019 (COVID-­19) vaccination: a U.S cohort. Obstet Gynecol.
2022;139:481–­9.
21. Edelman A, Boniface ER, Male V, Cameron ST, Benhar E, Han L, et al.
Association between menstrual cycle length and COVID-­19 vacci-
nation: global, retrospective cohort study of prospectively collected
data. BMJ Med. 2022;1(1):e000297.
22. Trogstad L. Increased occurrence of menstrual disturbances in 18-­
to 30-­year-­old women after COVID-­19 vaccination. SSRN Electron J.
2022; [cited 2022 Aug 1]. Available from: https://www.ssrn.com/abstr​
act=3998180 or http://dx.doi.org/10.2139/ssrn.3998180
23. Laganà AS, Veronesi G, Ghezzi F, Ferrario MM, Cromi A, Bizzarri
M, et al. Evaluation of menstrual irregularities after COVID-­19
vaccination: results of the MECOVAC survey. Open Med.
2022;17(1):475–­8 4.
24. Lee KM, Junkins EJ, Luo C, Fatima UA, Cox ML, Clancy KB.
Investigating trends in those who experience menstrual bleeding
changes after SARS-­CoV-­2 vaccination. Sci Adv. 2022;8:eabm7201.
25. Khan SM, Shilen A, Heslin KM, Ishimwe P, Allen AM, Jacobs ET,
et al. SARS-­CoV-­2 infection and subsequent changes in the men-
strual cycle among participants in the Arizona CoVHORT study. Am
J Obstet Gynecol. 2022;226(2):270–­3.
26. Li K, Chen G, Hou H, Liao Q, Chen J, Bai H, et al. Analysis of sex
hormones and menstruation in COVID-­19 women of child-­bearing
age. Reprod Biomed Online. 2021;42(1):260–­7.
27. Davis HE, Assaf GS, McCorkell L, Wei H, Low RJ, Re'em Y, et al.
Characterizing long COVID in an international cohort: 7 months of
symptoms and their impact. EClinicalMedicine. 2021;38:101019.
28. Tjaden AH, Fette LM, Edelstein SL, Gibbs M, Hinkelman AN,
Runyon M, et al. Self-­reported SARS-­CoV-­2 vaccination is consistent
with electronic health record data among the COVID-­19 community
research partnership. Vaccine. 2022;10(7):1016.
S U PP ORT I N G I N F OR M AT ION
Additional supporting information can be found online in
the Supporting Information section at the end of this article.
How to cite this article: Darney BG, Boniface ER,
Van Lamsweerde A, Han L, Matteson KA, Cameron
S, et al. Impact of coronavirus disease 2019
(COVID-­19) vaccination on menstrual bleeding
quantity: An observational cohort study. BJOG.
2023;00:1–10. https://doi.org/10.1111/1471-0528.17471